Is Autism Characterized by Enhanced Variability in Task-Related Brain Activation?

Background: Autism is characterized by a large diversity among symptomatic profiles and an important heterogeneity in individual developmental trajectories. One possible explanation for this variability in the autistic population is increased cerebral plasticity, the mechanism by which neural pathways can be modified by changes in neural processes. A wide range of studies has documented differences in autistics compared to typical ontrols in cerebral activation and performance, for example, during tasks involving visual and motor stimuli (Samson et al, 2011., Müller et al, 2004., Soulières et al, 2011). Enhanced cerebral plasticity is likely to be involved in these group differences, more specifically at the level of greater individual variability in the autistic group.

Objectives: The aim of the study is to use functional magnetic resonance imaging (fMRI) to determine whether there is enhanced variability in autism, in localization and intensity of cerebral activation within modalities that have already shown group differences.

Methods: Groups of 23 autistic and 22 typical participants matched on age (14-36 years old), FSIQ, Raven scores, and laterality performed a visuo-motor imitation task in an MRI scanner. Participants were instructed to imitate 96 visually presented hand gestures with the specified hand. SPM8 was used to study task-related brain activation in visual and motor regions of interest (ROI) at the group and individual level. For each participant the coordinates of the maximum activation peak were extracted in primary (Brodman area 4) and supplementary (Ba6) motor areas, primary (Ba17) and associative (Ba18+19) visual areas as well as in the fusiform gyrus (FG). An average coordinate for each group and for each ROI was computed by averaging all the individual stereotactic coordinates. For each ROI, each participant’s distance to his respective group average was then used as a variable of interest to determine group differences in variability. The same procedure was used to extract the beta value to evaluate differences in variability of intensity. A mixed-effects statistical model was then used to compare groups.

Results: Between group analysis showed more activation in autistics in associative visual areas (V4+Ba18) and the supplementary motor area, whereas controls showed more activity in parieto-occipital junction (V5) and extrastriate cortex (Ba19) (p<.05, FWE). Individual distances to the group mean were significantly higher in the autistic group than in the control group for the primary motor area (p=.01) and associative visual areas (p=.002). The opposite was observed for the primary visual area, where the distances were higher in the control than in the autistic group (p=.002). No between-group difference was observed for FG and for intensity of activation (p>.05).

Conclusions: Our results suggest that variability occurred in autism only in localization of activation, which was not generalized to all regions but was specific to some areas/modalities. The variability among autistics was not observed in some expected regions, e.g., the FG. Variability was not necessarily associated with regions in which differences in activation at the group level were observed. These results are consistent with an enhanced variability in the functional allocation of certain brain regions in autism.